PU box-binding transcription factors and a POU domain protein cooperate in the Epstein--Barr virus...

Downloaded from www.microbiologyresearch.org by

IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50

Journal of General Virology (1995), 76, 2679-2692. Printed in Great Britain 2679

PU box-binding transcription factors and a POU domain protein cooperate in the Epstein-Barr virus (EBV) nuclear antigen 2-induced transactivation of the EBV latent membrane protein 1 promoter

Anna Sj6blom, 1 Ann Jansson, 1 Weiwen Yang, ~ Sonia Lain, 2 Tina Nilsson ~ and Lars Rymo 1.

1 Department o f Clinical Chemistry and Transfusion Medicine, Sahlgrenska University Hospital and 2 Department of Medical Biochemistry, G6teborg University, S-413 45 Gothenburg, Sweden

Expression of the Epstein-Ban" virus (EBV) latent membrane protein (LMP1) is regulated by virus- and host cell-specific factors. The EBV nuclear antigen 2 (EBNA2) has been shown to transactivate a number of viral and cellular gene promoters including the promoter for the LMP1 gene. EBNA2 is targeted to at least some of these promoters by interacting with a cellular DNA binding protein, RBP-Jlc. In the present report we confirm and extend our previous observation that the LMP1 promoter can be activated by EBNA2 in the absence of the RBP-Jtc-binding sequence in the LMP1 promoter regulatory region (LRS). We show that two distinct LRS regions, - 106 to +40 and - 176 to - 136, contribute to EBNA2 responsiveness. Site-directed mutagenesis analysis of the upstream - 1 7 6 / - 1 3 6 EBNA2 responsive element revealed that two critical cis-acting elements are required for full promoter function. These same elements analysed by electro-

phoretic mobility shift assays define two binding sites recognized by nuclear factors derived from B cells. An octamer-like sequence ( - 1 4 7 to -139) contained overlapping binding sites for an unidentified transcriptional repressor on the one hand and a factor(s) belonging to the POU domain family but distinct from Oct-1 and Oct-2 on the other. An adjacent purine tract ( -171 to -155) held a PU.1 binding site, which was also recognized by a related factor. The results suggest that the POU domain protein and either of two PU box- binding factors bind simultaneously to LRS, creating a ternary complex that might be in part responsible for mediating the transactivation of the LMP1 promoter by EBNA2. There were no qualitative differences between EBV-negative and EBV-positive cells with regard to transcription factor binding to the octamer-like sequence and the PU.1 recognition site, as revealed by electrophoretic mobility shift assays.

Introduction

Epstein-Barr virus (EBV) is an ubiquitous human pathogen that has a strong association with at least three forms of cancer: African Burkitt's lymphoma (BL), nasopharyngeal carcinoma and a subset of Hodgkin's lymphomas (Henle & Henle, 1966; Herbst et al., 1991; Young et al., 1989). In addition, states of immune dysfunction, e.g. immunosuppression after organ trans- plantation, human immunodeficiency virus infection and primary immunodeficiencies may result in EBV-pro- voked lymphoproliferative disorders. EBV infection of human B lymphocytes in vitro leads to transformation and the outgrowth of EBV-carrying lymphoblastoid cell lines (LCLs) in which the virus genome is maintained in

* Author for correspondence. Fax +46 31 828458. e-mail [email protected]

multiple copies as autonomously replicating circular episomes (for reviews see Liebowitz & Kieff, 1993; Miller, 1990; and citations therein). Of the more than 80 genes encoded by the EBV genome, only a limited number are consistently expressed in LCLs: a family of six nuclear proteins, EBNA1 to -6; three membrane proteins, LMP1, -2A and -2B; and two small RNA molecules, EBER1 and EBER2. Six of these genes, EBNA1, -2, -3, -5 and -6 and LMP1 have been shown to be necessary for EBV-induced B lymphocyte growth transformation, while the others seem to be dispensible (Skare et al., 1985; Yates et al., 1985; Cohen et al., 1989; Hammerschmidt & Sugden, 1989; Mannick et al., 1991 ; Swaminathan et al., 1991 ; Longnecker et al., 1992, 1993; Tomkinson & Kieff, 1992; Kaye et al., 1993; Tomkinson et al., 1993).

EBNA2 is the first EBV-encoded protein expressed in infected B lymphocytes and plays an important role in the immortalization process. The results of Woiset- schlaeger et al. (1991) suggest that the EBNA2 message

0001-3308 © 1995 SGM


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50

2680 A. S jgblom and others

is transcribed from the Wp promoter at the initial phase of infection but within a short period of time there is a switch to a preferential use of the Cp promoter. The molecular mechanism underlying the switch is still unclear but the presence of EBNA2-responsive enhancer elements in the Cp regulatory region suggests that EBNA2 contributes either directly or indirectly (Woiset- schlaeger et al., 1990; Sung et al., 1991). A number of observations support the notion that EBNA2 exercises its effect on the phenotype of the EBV-infected cell by being a part of or by modulating the activity of regulatory systems that control the expression of specific viral and cellular genes. EBNA2 stimulates transcription from Cp and of the LMP1, LMP2A and LMP2B EBV genes (Abbot et al., 1990; FShraeus et al., 1990; Wang et al., 1990a; Zimber-Strobl et al., 1993; Sung et al., 1991; Woisetschlaeger et al., 1991). It activates the promoters for the CD21, CD23, c-fgr and c-bcl-2 genes (Cordier et al., 1990; Wang et al., 1987, 1990b, 1991; Knutson, 1990; Finke et al., 1992).

All EBNA2-inducible promoters characterized so far contain a 5' GTGGGAA 3' motif which is part of the recognition sequence for the cellular DNA-binding protein RBP-JK (Tun et al., 1994). It has been shown that RBP-JK binds to this motif in the context of EBNA2- responsive elements (EB2REs) and that RBP-JK interacts with EBNA2 both in the presence and absence of the DNA binding site (Ling et al., 1993 ; Zimber-Strobl et al., 1993; Henkel et al., 1994; Grossman et al., 1994; Yalamanchili et al., 1994; Laux et aI., 1994a; Waltzer et al., 1994; Johannsen et al., 1995). Thus, it remains an attractive hypothesis that RBP-Jtc may target EBNA2 to EB2REs. Recently, the macrophage- and B cell-specific transcription factor PU. 1 (Klemsz et al., 1990; Goebl, 1990; Ray et al., 1992) was implicated in the EBNA2- mediated transactivation of LMP1 (Johannsen et al., 1995) and the bidirectional LMP 1/TP2 promoters (Laux et at., 1994a). PU.1 was shown to bind to EB2RE in the promoter region and mutation of the recognition sequence completely abolished EBNA2 responses. Another member of the ets gene family of transcription factors, Spi-B, also bound to the same site (Laux et al., 1994b).

We have continued our investigation of the molecular mechanism by which EBNA2 activates the expression of the LMP1 gene. LMP1 is encoded by the BNLF1 reading frame transcribed in a leftward direction from a promoter at position 169546 in the EBV B95-8 genome (Baer et al., 1984; Fennewald et al., 1984). We have previously shown that the LMP1 promoter is controlled by positive and negative cis-elements in the Y-flanking region of the promoter and that EBNA2 can activate the promoter by overriding the effect of the putative repressors (F~hraeus et al., 1990, 1993). In the present

report we confirm and extend our previous observation that the LMP1 promoter can be activated by EBNA2 in the absence of the RBP-JK-binding sequence in the LMP1 promoter regulatory region (LRS). We show that two distinct LRS regions, - 106 to +40 and -176 to -136, contribute to EBNA2 responsiveness. Within the upstream - 1 7 6 / - 136 EBNA2 responsive sequence two cis-acting DNA elements, recognized by PU.I and a related factor and a POU domain protein, respectively, are shown to be necessary in conferring optimum promoter activity. Our results suggest that the POU domain protein may assist in the targeting of EBNA2 to the LMP1 promoter and that the POU domain protein and the PU box-binding factors cooperate in the transactivation of the promoter by EBNA2.

Methods Plasmid construction. All constructs were verified by dideoxy-

nucleotide chain termination sequencing utilizing the Sequenase system (USB). The pEAA6, pgSVECAT, pgCAT, p g L R S ( - 5 4 ) C A T , p g L R S ( - 1 0 6 ) C A T , p g L R S ( - 1 4 6 ) C A T , p g L R S ( - 2 1 4 ) C A T and p L R S ( - 106)CAT constructs have been described earlier (Ricksten et al., 1987; Ffihraeus et al., 1990, 1993). The LMP1 regulatory sequence is defined as nucleotides 169477 to 170151 of B95-8 EBV D N A (LRS), which corresponds to positions - 6 3 4 to +40 relative to the transcription initiation site. We would like to point out that the EBV sequence in the p g L R S ( - 146)CAT plasmid is identical to that of the p g L R S ( - 1 4 4 ) C A T plasmid in our previous report (Ffihraeus et al., 1993), in which 2 bp contributed to the EBV sequence by the linker went unnoticed. To obtain a true - 1 4 4 construct the pgLRS ( - 214)CAT plasmid was cleaved with Hind l I I and Nla l l l , overhanging 3' ends were removed with T4 D N A polymerase and the appropriate fragment was cloned in the pGem-3Zf (+ ) vector (Promega), resulting in the plasmid p g L R S ( - 1 4 4 ) C A T . The p g L R S ( - 1 4 6 ) ( - 2 1 4 / - 145)CAT plasmid was made by cloning three sets of double-stranded oligonucleotides with SalI ends in the SalI site of p g L R S ( - 146)CAT.

The pgLRS( - 106)(-- 1 7 0 / - 131)CAT plasmid was constructed by directional cloning of ligated sets of synthetic oligonucleotides, corresponding to the - 170 to - 131 region of LRS, into the PstI SalI site upstream of the - 1 0 6 / + 4 0 sequence in the p g L R S ( - 1 0 6 ) C A T construct. A series of mutated derivatives of the plasmid was constructed using synthetic oligonucleotides in which consecutive 5 bp segments of the normal - 1 7 0 to - 1 3 1 B95-8 EBV sequence were replaced by segments of pnrine-pyrimidine transversions: - 1 3 5 to - 1 3 1 ( m l ) , - 140 to - 1 3 6 ( m 2 ) , - 145 to - 1 4 1 ( m 3 ) , - 150 to - 146 (m4), - 155 to - 151 (m5), - 160 to - 156 (m6), - 165 to - 161 (m7) and - 170 to - 166 (m8), respectively. Similarly, the pLRS( - 106)( - 181 / - 145)CAT plasmid was constructed by cloning ligated sets of synthetic oligonucleotides corresponding to the - 181 to - 1 4 5 region of LRS into the SalI site in the p L R S ( - 1 0 6 ) C A T construct. Muta t ions were introduced as above, as 5 bp segments of purine-pyrimidine transversions in the following positions: - 1 4 9 to - 1 4 5 ( m l ) , - 155 to - 1 5 1 ( m 2 ) , - 160 to - 1 5 6 ( m 3 ) , - - 165 to - - 161 (m4), --170 to --166 (m5), - 1 7 5 to - 1 7 1 (m6) and - 1 8 1 to - 1 7 6 (m7), respectively.

The herpes simplex virus thymidine kinase (TK) promoter region in our T K constructs correspond to nucleotides - 108 to + 51 relative to the cap site (Edlund et al., 1985). The T K promoter-containing B a m H I fragment was excised from the p T K C A T plasmid (Ricksten et al., 1988) and inserted in a pGEM-3Zf( + ) vector in which the HindIII site


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50

E B N A 2 transact ivat ion o f the L M P 1 p r o m o t e r 2681

had been changed to a BamHI site, resulting in the pgTKCAT plasmid. LRS-containing derivatives were constructed by cloning sets of synthetic oligonucleotides corresponding to selected regions of LRS in the SalI site [pgLRS(-181/-145)TKCAT and pgLRS(-214/ -145)TKCAT] or between the PstI and Sall sites [pgLRS(-160/ -136)TKCAT, pgLRS(-194/-136)TKCAT and pgLRS(-217/ -136)TKCAT] upstream of the TK promoter in the pgTKCAT plasmid.

To make a series of mutated reporter plasmids with deletions covering the -176 to -147 region of LRS, PCR amplifications were performed using the pgLRS(-217)CAT plasmid as a template and primers that resulted in fragments with one end corresponding to position +40 in LRS and the other end corresponding to positions - 147, - 148, - 152, - 153, - 160 or - 176. The PCR fragments were cloned into the TA cloning vector (Invitrogen). Taking advantage of a synthetic HindlII site in one primer and a PstI site in the TA cloning vector, the PCR fragments were then cloned between the HindIII and PstI sites in the pgCAT plasmid. To generate the pgLRS(-217)CAT plasmid a set of four double-stranded synthetic oligonucleotides homologous to the -217 to - 7 4 region of LRS with overlapping single-stranded ends was annealed, creating an XmaI-PstI fragment. This DNA fragment and an isolated HindIII XmaI fragment of LRS corresponding to the -73 to +40 region were cloned between the HindIII and PstI sites of the pgCAT plasmid to create a continuous -217 to +40 LRS sequence in the plasmid.

Cell culture, DNA transfections and CAT assays. DG75 is an EBV genome-negative BL cell line (Ben-Bassat et al., 1977). The IB4 cell line was derived by transforming human placental lymphocytes with the B95-8 EBV strain (King et al., 1980). Cherry is an EBV-infected cell line that was established by culture of lymphocytes from patients with mononucleosis (Heller et al., 1981). Rael (Klein et al., 1972) and Raji (Epstein et al., 1966) are EBV-positive BL lines. The CBC-Rael line was obtained by infection of cord blood cells with the Rael virus strain (Ernberg et aL, 1989). The lymphoid cells were maintained as suspension cultures in RPMI 1640 medium (Life Technologies) supplemented with 10 % fetal calf serum (Life Technologies), penicillin and streptomycin. DG75 cells (5 x 106) were transfected with 10 lag DNA of the reporter construct to be tested and 800 fmol DNA of the EBNA2 expression vector pEAA6 or 800 fmol DNA of the pSV2gpt vector using the DEAE-Dextran technique (Ricksten et al., 1988). Cells were harvested after 48 h and aliquots of the cell lysates were assayed for CAT activity (Ricksten et al., 1988).

Electrophoretic mobility shift assays (EMSAs). Nuclear extracts were prepared as described by Dignam et al. (1983). The protease inhibitors antipain (5 gg/ml), leupeptin (5 lag/ml) and aprotinin (2 gg/ml) were added to the buffer in the final homogenization and dialysis steps. Aliquots were frozen in liquid nitrogen and stored at - 7 0 °C. EMSAs were performed using a double-stranded synthetic oligonucleotide corresponding to the - 173 to - 136 segment and - 176 to - 136 segment of LRS with single-stranded ends. The oligonucleotide was labelled by a repair reaction in the presence of [~-a~P]dNTP (6000 Ci/mmol; Du Pont NEN) using the Klenow fragment of DNA polymerase I (Boehringer Mannheim). Other probes used from the LRS region were - 153 / - 114, - 146 / - 114 and - 144 / - 114. The blunt-ended double-stranded oligonucleotides were labelled with [7- a2P]ATP (6000 Ci/mmol; Du Pont NEN) using polynucleotide kinase (Boehringer Mannheim). The labelled oligonucleotides were separated from free isotope by electrophoresis in a 4% polyacrylamide gel (acrylamide:bisacrylamide 30: 1) in 25 mM-Tris-HC1, 190 mM-glycine and 1 mM-EDTA pH 8.3. The wet gel was autoradiographed and the DNA fragments were excised, electroeluted by isotachophoresis (Ofverstedt et al., 1984) and precipitated with ethanol. Binding reactions (25 lal) contained 10 mM-Tris-HC1 pH 7.5, 50 mM-NaC1, 1 mM-DTT, 1 mM-EDTA, 5 % glycerol, 4 gg poly(dI-dC), 6 fmol 32p_

labelled DNA (approximately 70000 c.p.m.) and 20 lag of nuclear proteins (always added last). A 400-fold excess of competing oligonucleotide added before the 3~P-labelled probe was used for the competition experiments. After incubation on ice for 25 min, the samples were separated by electrophoresis on 5 % polyacrylamide gels (acrylamide: bisacrylamide 30: 1) in 25 mM-Tris-HCl, 190 mM-glycine and 1 mM-EDTA pH 8.3 for 3 h at 300 V.

Rabbit polyclonal antibodies against the transcription factors Oct-l, Oc~-2, PU.1/Spi-1 and Spl were purchased from Santa Cruz Biotechnology. The rabbit polyclonal antibody specific for POU domain proteins was raised against the POU domain of Oct-1 and was kindly provided by Dr Peter O'Hare (Marie Curie Research Institute, Oxted, UK). The anti-Spl antibody was used as a negative control in the PU. 1 supershift experiment. The supershift analyses were performed as described above for the binding experiments except that after the incubation on ice, 2 gl of the respective antibody was added. The mixture was incubated at 4 °C for 60 min and then analysed on 5 % polyacrylamide gels.

For in vitro expression, a fragment of pEAA6 containing the EBNA2-encoding open reading frame, BYRF1, was cloned into the plBI 31 vector (International Biotechnologies). The supercoiled DNA template was sequentially transcribed and translated in the same reaction mixture containing TNT rabbit reticulocyte lysate, amino acids and the TNT T7 RNA polymerase, using the protocol described by the manufacturer (Promega). Translated proteins were analysed by SDS- PAGE. The EMSA binding reactions were performed as described above except that 5 gl of the reticulocyte lysate was added together with the DG75 nuclear extract.

Results

Sequences in the - 2 1 4 / + 40 par t o f L R S responsible

f o r E B N A 2 responsiveness

W e h a v e p r e v i o u s l y d e m o n s t r a t e d t h a t E B N A 2 can

t r a n s a c t i v a t e the L M P 1 p r o m o t e r t h r o u g h sequences

d o w n s t r e a m o f p o s i t i o n - 2 1 4 in L R S (FAhraeus et al.,

1990, 1993; S j 6 b l o m et al., 1993). In o r d e r to m a p the

E B N A 2 - r e s p o n s i v e sequences in the - 2 1 4 / + 40 p a r t o f

L R S fur ther , a series o f C A T r e p o r t e r p l a s m i d s c o n -

t a in ing 5' d e l e t i o n m u t a t i o n s o f L R S was c o n s t r u c t e d

a n d sub jec t ed to the E B N A 2 c o t r a n s f e c t i o n assay in

D G 7 5 cells. T h e p l a s m i d tha t c o n t a i n e d o n l y the - 54 to

+ 40 p a r t o f L R S s h o w e d E B N A 2 i n d e p e n d e n t ac t iv i ty

(Fig. 1). A d d i t i o n o f the sequences b e t w e e n - 5 4 a n d

- 1 0 6 resu l ted in a s even fo ld a c t i v a t i o n o f the r e p o r t e r

p l a s m i d by E B N A 2 . T h e E B N A 2 re spons iveness was n o t

i nc reased fu r t he r by the a d d i t i o n o f the s e q u e n c e b e t w e e n

- 1 0 6 and - 1 4 4 . H o w e v e r , w h e n an a d d i t i o n a l 2 bp

( - 1 4 5 a n d - 1 4 6 ) were i n c l u d e d in the cons t ruc t , the

E B N A 2 re spons iveness was c o m p l e t e l y abo l i shed . A

p l a s m i d wi th an u n i n t e r r u p t e d L R S s e q u e n c e up to

p o s i t i o n - 2 1 4 s h o w e d a 47- fo ld a c t i v a t i o n by E B N A 2 .

I n s e r t i o n o f a 1 4 b p l inker at p o s i t i o n - 1 4 6 in a

c o n s t r u c t o t h e r w i s e iden t i ca l to p g L R S ( - 2 1 4 ) C A T

c o m p l e t e l y a b r o g a t e d the E B N A 2 r e spons ivenes s o f the

p l a smid . W e c o n c l u d e f r o m these e x p e r i m e n t s t ha t a t

least two sepa ra t e d o m a i n s o f the - 2 1 4 / + 4 0 p a r t o f

L R S p a r t i c i p a t e in the E B N A 2 - d e p e n d e n t a c t i v a t i o n o f


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50

2682 A. Sjdblom and others

+40 pgLRS(-54)CAT I

pgLRS( 106)CAT

pgLRS( 144)CAT I

pgLRS( t46)CAT [

pgLRS(-214)CAT I

pgLRS( 146)(-214~145)CAT I

-54 I

-106 I

144 I

CAT activity (%) Inducibility

-EBNA2 +EBNA2

3.1 (0.23) 7-1 (2.1) 2.3 (0-70)

0.55 (0.05) 3.8 (1.0) 7.1 (1.9)

pgCAT

146 I

145

II -146

-214 l

-214 I

0.42 (0.05) 216 (0144) 6.4 (1.4)

0-25 (0-04) 0-32 (0-08) 1-2 (0-I4)

0-43(0.04) 19 (2.1) 47 (7.8)

0.27 (0-04) 0.25 (0.02) 0.95 (0.04)

0-25 (0-04) 0-42 (0.01) 1-3 (0.22)

Fig. 1. Identification of two regions in LRS that contribute to EBNA2 responsiveness in DG75 cells. CAT reporter plasmids containing Y-deleted fragments of LRS were cotransfected with the EBNA2 expression vector pEAA6 ( + EBNA2), or an equivalent amount of the pSV2gpt plasmid ( - E B N A 2 ) into DG75 cells. The CAT activity is given as percentage chloramphenicol acetylation. Inducibility is defined as the ratio between the values obtained with the respective plasmids in the presence and in the absence of pEAA6. The value shown is the mean of four transfections and the SEM values are indicated within parentheses.

-140 -150 -160 170 -180 I t I I I

GCGGTGTGTGTGTGCATGTAAGCGTAGAAAGGGGAAGTAGAAAGCGTGTQT

octamer PU box 106 + D~ - > < Dfl > 5.0

pgLRS(- 106)(- 170/- 131)CAT ---t p '

Mutation: ml " ' - t ~

m2 "'-"1

m3 ----t t::::::::::::::~

m4 -{ ~:::-::::::::::r

m5 ---I t::::::::::::::l

m7 -.-] ~::::::::::::::~--

m8 ---~ ~::::::::::::::1

pgLRS( 106)CAT--- I

-106 pLRS(-106)(- 181/-145)CAT ---1

Mutation: ml " ' 4 "~:::'::::::::::J

m2 . - 4 r:::::::::::::J

m3 ---I I:.:-:.::;::::1

m4 -- --1 t:::::,:::::q

m5 .---I t:::::::::,::::,

m6 ----I i:.:.:+:+:~

m7 "'--I t:.:.:.:.:-:-:+--

pLRS(-106)CAT ----I

CAT activity (%) Inducibility 10 15 20 25 30

i t

35 40

I

I i

I

f

l

% ,

I - E B N A 2

O+EBNA2

50 (11)

40 (10)

29 (11)

19 (5-8)

14 (2-6)

27 (3-2)

15 (3.6)

14 (2.5)

10 (4.2)

8.2 (1.3)

6-9 (0.65)

6.5 (0.49)

6-6 (1-2)

7.8 (1.4)

6.8 (1-6)

8.5 (2-2)

10 (1-9)

8.9 (1-5)

5.6 (1-6)

Fig. 2. Mutat ional analysis of EBNA2-responsive elements in the - - 1 8 1 / - 1 3 1 region of LRS. Stretches of purine-pyrimidine transversions were introduced into the p g L R S ( - 1 0 6 ) ( - 1 7 0 / - 1 3 1 ) C A T and p L R S ( - 1 0 6 ) ( - 1 8 1 / - 1 4 5 ) C A T plasmids at positions indicated by dotted boxes using synthetic oligonucleotides (see Methods). The reporter plasmids were cotransfected with pEAA6 ( + EBNA2) or with an equivalent amount of pSV2gpt ( - E B N A 2 ) into DG75 cells. The CAT activity is given as percentage chloramphenicol acetylation. Inducibility is defined as the ratio between the values obtained with the respective plasmids in the presence and in the absence of pEAA6. The values shown are the mean of four transfections. The SEM values are indicated with error bars for CAT activity and given as numbers within parentheses for inducibility by EBNA2.


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50

EBNA2 transactivation of the LMP1 promoter 2683

Table 1. Analysis of EBNA2-dependent activation of the TK promoter mediated by the - 2 1 7 / - 131 LRS region

CAT activity (%)*

Reporter plasmid - EBNA2 + EBNA2 Inducibilityt

pgLRS(- 217/- 136)TKCAT pgLRS(- 214/- 145)TKCAT pgLRS( - 194/- 136)TKCAT pgLRS(- 181/- 145)TKCAT pgL RS( - 160/- 136)TKCAT pgTKCAT

2.4 (0.30) 33 (3.0) 15 (3.1) 2-0 (0.17) 7.7 (0.58) 3.9 (0.17) 1.6 (0.18) 15 (1.1) 10 (1.6) 2.0 (0.28) 5.2 (0.43) 2-7 (0.26) 1-2 (0-18) II (0-73) %6 (1.8) 2.1 (0.17) 4.7 (0.26) 2-3 (0.20)

* Expressed as percentage acetylation of chloramphenicol. The values shown are the mean of four transfections; the SEM is given in parentheses.

t EBNA2 inducibility is calculated as the ratio between the CAT value obtained with and without EBNA2. The SEM is given in parentheses.

the LMP1 promoter. One of the domains is located upstream of position - 1 4 6 and the other closer to the promoter between position - 1 0 6 and - 5 4 . The - 1 4 6 / - 5 4 region seems to contain two or more negative cis-elements with the ability to repress the constitutive activity of the promoter-proximal elements. Interestingly, the insertion mutation only interfered with the activating but not the suppressing functions of LRS and prevented both upstream and downstream positive elements from activating the LMP1 promoter.

The results of preliminary mutational analysis (data not shown) and DNase I footprinting experiments (Sj6blom et al., 1993) suggested that the element(s) responsible for the major part of the EBNA2-induced transactivation found above and around position - 146 of LRS were contained within the - 176 to - 136 region. In order to define these regulatory elements in greater detail, 5 bp stretches of purine-pyrimidine transversions were introduced in this region, resulting in two different series of mutated reporter plasmids, derived from p g L R S ( - 106)(- 1 7 0 / - 131)CAT and p L R S ( - 106)- ( - 1 8 1 / - 1 4 5 ) C A T . The plasmids were cotransfected with the EBNA2 expression vector in DG75 cells and analysed for CAT activity. The results obtained with the p g L R S ( - 106)(- 1 7 0 / - 131)CAT series revealed the presence of at least two LRS subdomains important for EBNA2 responsiveness (Fig. 2). For simplicity, the one located at position - 1 5 0 / - 1 4 1 was designated De and the one at position - 1 7 0 / - 156 was designated Dfl. De contains a sequence with partial identity to the octamer motif and Dfl contains a purine-rich sequence resembling a PU box. It should be noted that although the level of activity was higher, EBNA2 inducibility (defined as the ratio between the activity obtained in the presence and in the absence of EBNA2) of the p L R S ( - 1 0 6 ) ( - 1 8 1 / - 145)CAT plasmid containing only the Dfi region was similar to that of the p L R S ( - 1 0 6 ) C A T plasmid. This was also true for the

mutated plasmids belonging to this series due to the fact that the mutations induced a proportional decrease of the activity, both in the presence and absence of EBNA2. In contrast, the EBNA2 inducibility of the p g L R S ( - 106)(- 1 7 0 / - 131)CAT plasmid, which contains both the De and Dfl regions, was increased about sixfold compared with p g L R S ( - 106)CAT. Mutations in either De or Dflin the p g L R S ( - 106)(- 1 7 0 / - 131)CAT series not only reduced total activity but also the inducibility of the plasmids. The Dfl mutations reduced the activity to the background [the p g L R S ( - 1 0 6 ) C A T level]. We conclude from these experiments that (i) Dfl by itself functions as a positive transcription element that is not inducible by EBNA2; (ii) De is not active in the absence of Dfl; and (iii) elements in De and Dfl cooperate to create a fully EBNA2-responsive site.

To investigate whether the putative EBNA2-responsive element in the - 2 1 4 to - 1 3 6 region could confer EBNA2 inducibility to a heterologous promoter in the absence of the downstream EBNA2-responsive LRS elements, fragments of this region were cloned in front of the TK promoter in a reporter plasmid (Table 1). The results showed that the - 2 1 7 / - 136 LRS region indeed conferred EBNA2 responsiveness to the TK promoter- carrying construct [pgLRS( - 2 1 7 / - 136)TKCAT]. De- letion of the - 1 4 4 / - 1 3 6 part of LRS resulted in constructs [e.g. p g L R S ( - 2 1 4 / - 145)TKCAT] with only about 20 % of the original activity, suggesting that this sequence constitutes an important part of the EBNA2- responsive element.

Detailed analysis of the De domain of LRS

The reason for the complete loss of EBNA2 responsiveness of the p g L R S ( - 146)CAT construct is not clear. To clarify precisely at what point the repression was relieved and the EBNA2 inducibility restored, a fine mapping of the - 1 7 6 / - 144 LRS region was carried out


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50

2684 A. Sj6blom and others

EBNA2: + + + + + + + + + + + +

TLC ~I~ ~ ~

CAT activity 47 11 6.4 0.3 4-4 4,8 6.6 60 13 68 59 0.6 >100 (% acetylation)

Fig. 3. Deletion analysis of the octamer-homologous region of LRS for EBNA2 responsiveness. Reporter plasmids with deletions covering the - 217 to - 54 region of LRS (see Methods) were cotransfected with the EBNA2 expression vector pEAA6 (+ EBNA2) into the EBV-negative DG75 cell line. The CAT activity is presented as percentage chloramphenicol acetylation. The experiment shown is a representative example out of five transfections.

by 5' deletion mutat ion analysis (Fig. 3). As expected, the - 146 LRS construct did not respond to EBNA2 in the EBNA2 cotransfection assay in DG75 cells. Deletion of 2 b p [pgLRS(-144)CAT] or the addition of 1 bp [ p g L R S ( - 147)CAT] restored EBNA2 responsiveness to

the level obtained with p g L R S ( - 1 0 6 ) C A T . It might be of significance that a complete octamer-homologous mot i f was present in the p g L R S ( - 1 4 7 ) C A T plasmid. Further addition of LRS sequence up to - 160 had only minor effects on activity. However, when the complete sequence downstream of position - 176 was included in the construct a strong increase in responsiveness was obtained. Addition of the sequence between - 1 7 6 and - 2 1 7 had only a minor effect.

These results are consistent with the hypothesis that a trans-acting factor with the ability to block the effect of all positive elements in the - 2 1 4 / + 40 part of LRS binds close to position - 146 in p g L R S ( - 146)CAT. The effect of this putative repressor is eliminated in the p g L R S ( - 1 4 4 ) C A T and p g L R S ( - 1 4 7 ) C A T plasmids. To correlate the activity data with the potential binding of transcription factors we performed EMSAs with nuclear extracts of EBV-negative DG75 and EBV- positive Cherry cells. Radioactively labelled, double- stranded synthetic oligonucleotides corresponding to the - 1 5 3 / - 1 1 4 , - 1 4 6 / - 1 1 4 and - 1 4 4 / - 1 1 4 LRS regions were used as probes (Fig. 4). Three specific complexes were recognized both in EBV-negative and EBV-positive cells and localized to specific regions of LRS

(a) ,v \ .v ~ .@ ,~> > >

), Y Y ) , Y y

(b)

S I E ~ ~ S I E ~

D~2=~ ~ .... D ~ I ~ D ~ 2 ~

LRS(-146/ 114)

(c)

SIE

Dc~l

LRS(-153/-114)

I I I I F .1~ I F

DG75 cells Cherry cells DG75 cells Cherry cells DG75 cells Cherry cells

Antibody Antibody OCT-1 - + + OCT-1 + + OCT-2 + + - OCT-2 + + P OU + + POU + +

Fig. 4. Analysis of transcription factors binding to the octamer motif-containing Dc~ region of LRS. EMSAs using nuclear extracts of the EBV-negative DG75 and EBV-positive Cherry cells were performed with the - 1 5 3 / - 1 1 4 , - 1 4 6 / - 1 1 4 and 1 4 4 / - 1 1 4 fragments of LRS as probes. (a) The EMSA pattern obtained with the three probes and B cell extracts. The specific complexes are indicated by black arrows and designated Dc~l, D~2 and SIE. A non-competable unspecific band is indicated by the hatched arrow. The fastest migrating band is the free probe. Autoradiograms (b) and (c) are the result of EMSA supershift experiments using the - 146/-- 114 and - 1 5 3 / - 114 LRS probes and rabbit polyclonal antibodies against the Oct-1 and Oct-2 factors and the POU domain.


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50


(a)

[Dcd, Dill/Dil2]

DcO D/~ Dill

LRS(-173/-136)

1 2 3 4 5 6 7 8 9 10 11 12 (b)

(I) -136 -140 -150 -160 -170 -173

I I I I I I

5 "-TCGACGTGTGTGTGCATGTAAGCGTAGAAAGC-W-_4~AAGTAGAAA- 3" 3 "-GCACACACACGTACATTCGCATCTTTCCCCTTCATCTTT- 5"

<-- D~ ---> < D~ >

(II) 5"-TCGACCACACACACGTACATTCGCATCTTTCCCCTTCATCTTT-3" 3" GGTGTGTGTGCATGTAAGCGTAGAAAGGGGAAGTAGAAA-5"

(]11) 5"-TCGACCACACACACGTACATTCGCATCTTTGGGGAAGTAGAAA-3" 3"-GGTGTGTGTGCATGTAAGCGTAGAAACCCCTTCATCTTT-5"

(w) 5"-GTGTGTCACGTTGTAAGCGTAGAAA~AAGTAGAAA-3" 3"-CAGTGCAACATTCGCATCTTTCCCCTTCATCTTT-5"

(v) 5"-GTGTGTGTGCATGTAAGCGTTCTTTCCCCTTCATCAAA-3" 3"-CACACGTACATTCGCAAGAAAGC-C4~AAGTAGTTT-5"

I I } t

DG75 cells IB4 cells

Mutat ion in competitor (I) Unmuta ted - + + - - - (II) -173/-136 + + (III) 160~136 + + (IV) 146/-142 + . . . . . + (V) -170/-156 +

Fig. 5. Binding of factors in B lymphoid cells to the Dc~/Dil region. (a) A 32P-labelled double-stranded synthetic oligonucleotide corresponding to the - 1 7 3 / - 136 LRS region was incubated with nuclear extracts from DG75 (lanes 1 6) and IB4 (lanes 7-12) cells and subjected to EMSA. Lanes 1 and 7 in the autoradiogram show the binding pattern obtained with the nuclear extracts. Competi t ion reactions performed by adding a 400-fold excess of unlabelled competitor to the binding mixtures were analysed in the rest of the lanes using oligonucleotides I to V with mutat ions in specific regions as indicated below the autoradiogram. Four specific complexes are indicated by black arrows and designated Dc~ 1, Dill, Dfl2 and [D~ 1, Dill/Dil2]. Three non-competable, unspecific bands are indicated by hatched arrows. (b) Nucleotide sequence of the double-stranded oligonucleotides I to V used in the competition experiment. Mutated nucleotides in the oligonucleotides II-V are underlined.

by competition experiments (data not shown). Two of the EMSA bands were due to factors that interacted with the Dc~ region and were designated Dc~l and Da2 (Fig. 4). The third complex corresponded to a factor that interacted with the - 1 2 7 / - 1 1 9 LRS region, which contains a sequence (5' TTCCCGAAA Y) with partial identity to the binding site for the sis-inducible factor (SIF; Sadowski et al., 1993). This complex, designated SIE (sis-inducible element) in Fig. 4, was competed for with an oligonucleotide that contained a consensus SIE (Santa Cruz Biotechnology; data not shown). The prominent band between the De2 and SIE bands indicated with a broken arrow in Fig. 4 did not compete with an excess of unlabelled probe and was therefore judged as unspecific (data not shown). The D~2 complex was formed when the - 1 4 6 / - 114 probe was used (Fig. 4, lanes 2 and 5). Interestingly, this complex was more easily competed for with an oligonucleotide that ex- tended to - 153 than - 146, suggesting that nucleotides upstream of - 1 4 6 are involved in the protein binding (data not shown). The D~2 complex was, however, more or less completely replaced by the D~I complex in EMSAs with the - 1 5 3 / - 114 probe (Fig. 4, lanes 1 and 4). Neither the Del nor the Dcd complex was formed with the - 1 4 4 / - 114 probe (Fig. 4, lanes 3 and 6). We

conclude from these results that D~ contains two overlapping binding sites for the Del and D~2 factors with essential sequences located between position - 146 and - 1 4 4 for the D0~2 site and somewhat further upstream for the D~I site. Under our EMSA conditions, binding of the D~I factor was always dominant when the complete Dc~ region was present, both with EBV-negative and EBV-positive cell extracts. It should be noted that the binding of the D~2 factor to the - 1 4 6 / - 114 LRS probe correlated with the loss of EBNA2 responsiveness of the corresponding p g L R S ( - 146)CAT plasmid. This suggests that the D0d factor functions as a negative regulator of LRS activity. In contrast, binding of the D~I factor to the - 1 5 3 / - 1 1 4 probe was paralleled by a high EBNA2 responsiveness of the corresponding reporter plasmids, indicating the D~I factor is a transactivating protein.

Since the D~ region contained a sequence with a partial identity to the octamer motif we then asked if any of the D~-binding proteins belonged to the oct gene family of transcription factors. Supershift experiments were performed with the - 1 5 3 / - 114 and - 1 4 6 / - 114 LRS probes and DG75 and Cherry cell extracts, employing polyclonal antibodies against the Oct-1 and Oct-2 factors and the POU domain, respectively. The


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50


activity and specificity of the antibodies were verified in supershift experiments using a probe carrying the octamer consensus sequence and B cell extracts (DG75, Cherry, IB4; data not shown). The antibodies also reacted with Oct-1 and Oct-2 proteins translated in vitro (data not shown). In EMSAs with the LRS probes the anti-Oct-1 and -Oct-2 antibodies did not detectably shift the De l or the De2 EMSA band (Fig. 4). The anti-POU domain antibodies, on the other hand, very distinctly depleted the De l band, indicating that the corresponding protein belongs to the POU domain family of transcription factors but is different from Oct-1 and Oct-2. The De2 complex was not affected by the addition of antibodies. The De2 factor is therefore different from De l .

LRS(-176/-136)

Dcd, Dfll/Dfl2 ...~.

Dal

D/~2 D/~I

Factors belonging to the ets gene family of proteins interact with Dfl

Our results showed that elements in Dfl were necessary to achieve maximum activity of LRS-containing reporter plasmids in the presence of EBNA2 (Sj6blom et al., 1993). To identify factors that interact with the Dfl region, we performed EMSAs with nuclear extracts of DG75 and IB4 cells using a radioactively labelled, double-stranded synthetic oligonucleotide corresponding to the - 1 7 3 / - 1 3 6 region of LRS as a probe (Fig. 5a, b). There was no consistent difference between the EBV-negative and the EBV-positive cells with regard to the pattern of complexes. Competition experiments with unlabelled oligonucleotides I V (Fig. 5b) were carried out in 400-fold molar excess over the labelled fragment to define shorter sequences within the 37 bp LRS probe that were involved in protein-DNA interactions. The EMSAs revealed four specific complexes: D e l , Dill, Dfl2 and [Del , Dfll/Dfl2] (Fig. 5a; lanes 2 and 8 as compared to lanes 3 and 9). The Dill and Dfl2 bands had very similar mobility and were only resolved under ideal conditions. The weak bands indicated by broken arrows in Fig. 5 were not competed for and were assumed to represent unspecific complexes. Competition with an oligonucleotide containing a mutated sequence between position - 160 and - 136, which includes the De domain (Fig. 5a; lanes 4 and 10), or an oligonucleotide with a mutated octamer motif between nucleotides - 1 4 6 to - 1 4 2 (Fig. 5a; lanes 5 and 11) competed for all complexes except De l . Competition with an oligonucleotide that contained mutations covering the Dfl region between - 170 to - 156, competed for all but the Dill and Dfl2 complexes (Fig. 5a; lanes 6 and 12). Notably, mutations that interfered with the formation of either the D e l or Dfll/Dfl2 complexes also inhibited the formation of the [Del , Dfll/Dfl2] complex. The results suggest that DG75 and IB4 cells both contain a similar

DG75 cells Antibody

Oct- 1 - + Oct-2 + - POU - - - +

Fig. 6. Immunological analysis of factors in B cells that bind to the De domain. Nuclear extracts from DG75 cells were incubated under binding conditions with a 32P-labelled double-stranded oligonucleotide corresponding to the - 176/- 136 LRS region followed by incubation with rabbit polyclonal antibodies against Oct-l, Oct-2 and the POU domain as shown underneath. The reaction mixtures were analysed by EMSA. Four specific complexes are indicated by black arrows and designated D~I, Dill, Dfl2 and [Dc~l, Dfll/Dfl2]. A non-competable, unspecific band is indicated by the hatched arrow.

set of transcription factors with affinity for the De and Dfl domains. Furthermore, the properties of the [Del , Dfll/Dfl2] band suggest that it is formed by the simultaneous binding of factors to the De and Dfl sites and that the complex is labile so that it is barely detectable by EMSA analysis (see also Fig. 6 and 7).

Supershift experiments were performed with nuclear extracts of different B cell lines and the - 1 7 6 / - 1 3 6 LRS probe, using specific antibodies against the transcription factors Oct-l, Oct-2 and PU.1 and the POU domain protein family (Fig. 6; Fig. 7). The results confirmed our previous observation that a protein interacting with De (the D e l factor) was recognized by the anti-POU domain antibody. Furthermore, the [Del , Dfll/Dfl2] band was also depleted by this antibody, indicating that the complex contained a POU domain protein. None of the complexes reacted with the anti-


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50


LRS(-176/-136)

D~I

Dfl2 D/~I

, ~ ~ , ~ t Cherry~ ~ DG75 Rael CBC- IB Raji Antibody Rael

Control + + + - + - + - + - PU.1/Spi-1 - + - + + + + +

Fig. 7. Immunological analysis of factors that bind to the Dfl domain in different B cell lines. Nuclear extracts from DG75, Rael, CBC-Rael, IB4, Cherry and Raji cells were incubated under binding conditions with a 32P-labelled - 1 7 6 / - 136 LRS oligonucleotide probe and with control (anti-Spl) or anti-PU.1/Spi-I antibodies as indicated. The reaction mixtures were analysed by EMSA. Three specific complexes, D~I, Dill and Dfl2, are indicated by black arrows. A non-competable, unspecific band is indicated by the hatched arrow.

Oct-1 and -Oct-2 antibodies (Fig. 6). One of the factors that interacted with Dfl was recognized by the PU. 1- specific antibody, as shown by the elimination of the Dill complex in all cell lines tested (Fig. 7). Furthermore, an oligonucleotide that contained a consensus PU. 1-binding motif inhibited the formation of both the Dill and Dfl2 complex in an EMSA competition experiment, suggesting that the Dill and Dfl2 factors belong to the same family of DNA-binding proteins, the ets gene family (data not shown). The Dill and Dfl2 factors were present in both the EBV-negative and EBV-positive B cell lines, although the relative amounts varied between the lines (Fig. 7).

EBNA2 is targeted to the LMP1 promoter by the POU domain pro te in

Having shown that D~ can confer EBNA2 responsiveness to an heterologous promoter much more efficiently than Dfl, we asked whether it would be possible to demonstrate a direct interaction between

LRS~176/-136)

D~I

Dfl2 Dill

DG75 cells In vitro-translated EBNA 2 - + - Control - - +

Fig. 8. Effect of the addition of in vitro-translated EBNA2 on protein binding to the LRS probe. DG75 nuclear extracts were incubated with the 32P-labelled - 1 7 6 / - 1 3 6 LRS oligonucleotide probe and aliquots of reticulocyte in vitro translation reactions with control DNA or EBNA2 DNA as indicated. The binding mixtures were analysed by EMSA. The positions of the three previously identified complexes, D~I, Dill and Dfl2, are indicated by black arrows. A non-competable, unspecific band is indicated by the hatched arrow. The black arrow at the top indicates a complex which was induced or strongly enhanced by the addition of in vitro-translated EBNA2.

EBNA2 and factors bound to the D~ or Dfl region. Accordingly, we assayed binding of EBNA2 expressed in a reticulocyte lysate translation system to a - 1 7 6 / - 136 LRS probe by EMSA in the presence of transcription factors furnished by DG75 nuclear extract (Fig. 8). The presence of recombinant EBNA2 in the binding mixture resulted in a specific elimination of the D~I band corresponding to the POU domain protein, which was not seen when a reticulocyte lysate containing the vector


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50

2688 A. S jdbIom and others

control only was added. The PU. 1-related complexes in the Dfl region were not affected. The addition of the EBNA2 protein also resulted in an increase in mass in the EMSA pattern at a position corresponding to a slower migrating band (Fig. 8). Competition experiments showed that this complex was specific (data not shown), but the nature of its constituents is not clear. Our interpretation of these data is that EBNA2 can make a direct contact with the POU domain protein but the molecular consequences of the interaction remains unknown.

Discussion

It is now well established that a major way by which EBNA2 exerts its effect on the EBV-infected cell is by regulating the activity of certain viral and cellular promoters. We and others have previously reported that LRS contains c is -e lements that mediate the EBNA2- induced up-regulation of LMP1 expression (F~hraeus et al., 1990, 1993; Ghosh & Kieff, 1990). Our previous characterization of LRS indicated that the region between -214 and +40 is sufficient to direct EBNA2- induced transactivation of LMP1 promoter-containing plasmids in B cells. We now provide evidence that two different regions of this part of LRS, - 176 to - 136 and -106 to +40, contribute to EBNA2 responsiveness of reporter plasmids independently of each other. We have defined transcriptional elements and factors in the - 1 7 6 / - 1 3 6 region that participate in the EBNA2- induced transactivation of the LMP1 promoter.

The upstream region of the EBNA2-responsive LMP 1, LMP2A, Cp and CD23 promoters all contain the DNA sequence 5' GTGGGAA 3'. A number of recent reports demonstrate that this motif in the context of an EBV DNA promoter sequence specifically interacts with a DNA-binding cellular protein, RBP-J~c, that can target EBNA2 to its responsive element, and that this interaction results in promoter activation (Ling et al., 1993; Zimber-Strobl et al., 1993; Grossman et al., 1994; Henkel et al., 1994; Meitinger et al., 1994; Waltzer et al., 1994; Yalamanchili et al., 1994; Johannsen et al., 1995). The core sequence of the RBP-J~c-binding site is present at positions - 2 2 3 / - 2 1 7 and - 2 9 8 / - 2 9 2 in LRS, which were not included in the sequence analysed in this report. Our previous results using deletion analysis and site-directed mutagenesis have, however, very clearly demonstrated that EBNA2 can activate the LMP1 promoter in reporter constructs lacking the RBP-JK- binding site (F~hreaus et al., 1990, 1993; Sj6blom et al., 1995). This has recently been confirmed by Johannsen et al. (1995). It is also consistent with the observation that RBP-JK has a much lower affinity for its binding site in

the LRS context than for the corresponding sites in the Cp or CD23 promoters (Ling et al., 1994). On the other hand, it has been reported by Laux et al. (1994a) that LRS sequences between -232 and -199 are essential for EBNA2 responsiveness of the LMP1 promoter in B cells. The reason for this discrepancy remains unex- plained.

Our results show that the - 1 7 6 / - 1 3 6 LRS region contains two domains, here denoted De ( - 1 5 0 / - 1 4 1 ) and Dfl ( - 170/ - 156) (Fig. 2), of particular importance for EBNA2 responsiveness. Constructs that contained both De and Dp had a higher EBNA2 inducibility than the corresponding deleted plasmid [pgLRS(- 106)CAT] and mutations in either element reduced EBNA2 inducibility. In contrast, although the level of activity was higher in plasmids that contained only Dfl as compared with the deleted construct, EBNA2 inducibility was the same. This was true regardless of whether Dfl was mutated or not. Our interpretation of the results is that (i) elements in De and Dfl must cooperate to achieve maximum EBNA2-responsiveness of the - 1 7 6 / - 1 3 6 region; (ii) in the absence of De the Dfl element functions mainly as a positive element, ampli- fying the activity of elements in the - 1 0 6 / + 4 0 LRS region.

It should also be noted that the - 176 LRS construct had about the same EBNA2 inducibility as that which contained the complete region up to -217 (about 47- fold induction), showing that no essential elements were deleted (Fig. 3). Essentially the same result was obtained with a -217 LRS construct in which the - 2 1 7 / - 1 9 5 sequence had been mutated (Sj6blom et al., 1993). On the other hand, Johannsen et al. (1995) have reported that a -215 to +40 LMP1 promoter construct was EBNA2-responsive but a - 2 0 5 / + 40 construct was not. They also mapped binding sites for unidentified transcription factors to the - 2 1 5 / - 2 0 5 region (LBF3, 5, 6 and 7 in B cells) which were assumed to contribute to EBNA2 responsiveness (Johannsen et al., 1995). In iine with the latter observation, we have recently obtained evidence from experiments with deletion mutants of EBNA2 which suggest that part of the transactivating effect of EBNA2 might be mediated through sequences in the - 2 1 4 / - 195 LRS region (A. Sj6blom, A. Nerstedt, A. Jansson & L. Rymo, unpublished results). The reason for the inconsistency between these results is not clear. We are presently investigating if differences in reporter constructs, expression vectors, cell lines or transfection conditions may play a role. It is conceivable that the artificial rearrangement of promoter elements or the introduction of mutations associated with in vitro studies might have had consequences that do not properly reflect the function of the promoter in the native state.

Protein-binding studies employing EMSAs suggested


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50

E B N A 2 transact ivat ion o f the L M P 1 p r o m o t e r 2689

that both EBV-negative and EBV-positive cells contain apparently similar sets of transcription factors that bind to the D~/Dfl region. Furthermore, it was not possible to identify EBNA2 as a component of the EMSA complexes obtained with IB4 nuclear extracts by supershift experiments with monoclonal antibodies (data not shown). The only detectable difference with regard to protein- binding between B cells that contained EBNA2 and those which did not was a somewhat diminished protection of the Dc~ domain when studying the IB4 cell extract in the DNase I footprint analysis (Sj6blom et al., 1995). Consistent with this difference in the Dc~ domain, a somewhat changed intensity of the D~I complex was seen when compared in the EMSA studies using EBV- negative and EBV-positive EBNA2-containing cells. This might be an indication of a decreased stability or a conformational change of the complex between the elements in the Dc~ domain and cellular factors in EBNA2-containing cells.

The sequence at position - 1 4 7 / - 1 4 0 in Dc~ with partial identity (6/8) to the octamer motif suggested that a member of the Oct family of transcription factors might be responsible for the effect of the D0~ element on LMP1 promoter activity. We have, however, not been able to demonstrate the binding of the Oct-1 or Oct-2 factors, which are expressed in B cells, to this site with specific antibodies. The Oct factors belong to a larger family of transcription factors designated POU domain proteins (Wegner et al., 1993). These proteins share a conserved DNA-binding domain (the POU domain) consisting of the POU-specific 75 to 82 amino acid domain, a short variable linker region and the homeo- domain. Using an antibody against the POU-specific domain we were able to demonstrate that both EBV- positive and EBV-negative B cells contain a distinct factor that binds to the octamer motif in the context of surrounding LRS sequences. If sequences upstream of position - 146 in the binding site were deleted, the POU domain protein was replaced by an unidentified factor, suggesting that D~ contains two overlapping protein- binding sites. Interestingly, the binding of the unidentified factor correlated with a complete loss of EBNA2 responsiveness. Deletion of a further 2 bp resulted in the loss of binding of the unidentified factor, which correlated with the restitution of EBNA2 responsiveness to almost the same level as when the POU domain protein was bound (provided no elements upstream of D~ were present in the promoter constructs). It is, however, not immediately obvious what role this putative silencer might play in the regulation of LMP 1 promoter activity, since the binding of the POU domain protein to a DNA fragment that contained an unmutated D~ always dominated over that of the unidentified factor both in EBV-positive and EBV-negative B cells. It is of course

possible that our assay conditions do not properly reflect the in vivo situation.

It has recently been reported that the PU.1 transcription factor binds to the PU box present in Dfl (Laux et al., 1994b; Johannsen et al., 1995). PU.1 is expressed specifically in haematopoietic tissues, particularly in the monocytic and B lymphoid lineages (Klemsz et al., 1990; Goebl, 1990; Ray et al., 1992). Our protein-binding studies showed that two discrete factors bound to an element in the DE domain (the D/~I and Dp2 complexes in Figs 5 and 7) and competition experiments with an excess of a PU box consensus oligonucleotide resulted in the removal of both complexes (data not shown). EMSA supershift experiments using an anti-PU.1 antibody depleted the complex corresponding to the Dill band in all B cell lines tested, confirming the conclusions of Laux et al. (1994b) and Johannsen et al. (1995). In addition, Laux and colleagues showed that a second member of the Ets family of transcription factors, Spi-B, also recognized the same PU.1 recognition site (Laux et al., 1994b). Thus, it seems reasonable to assume that the D~2 complex in our study is formed by the binding of Spi-B to DE. Spi-B is expressed in all human haematopoetic cell lineages except T cells and has been shown to transactivate reporter plasmids containing PU boxes (Ray et al., 1992).

EMSAs revealed that a relatively abundant factor in B cells bound to the - 1 2 7 / - 1 1 8 LRS region which contains a sequence with partial identity to the binding site for SIF. An oligonucleotide that contained a consensus SIE inhibited the binding. SIF is a DNA- binding protein that is induced by polypeptide growth factors and the presence of an SIE in a promoter region is sufficient under certain circumstances to mediate growth factor-activated transcription (Sadowski et al., 1993). However, this factor did not seem to be essential for the regulation of LMP1 promoter activity, at least not in the context of the - 1 4 4 / + 4 0 LRS region, since deletion of SIE had little effect on EBNA2 responsiveness [compare pgLRS(-144)CAT and pgLRS(-106)CAT constructs in Fig. 1].

How does EBNA2 contact LRS? Evidence accumu- lating from several groups strongly indicates that the EBNA2 protein does not bind directly to a specific DNA sequence but acts via protein-protein interactions with the LMP1 promoter regulatory region. EBNA2 might also act at a distance. We have recently reported evidence suggesting that one of the functions of EBNA2 might be to inhibit dephosphorylation by protein phosphatase 1 leading to a modulation of the activity of transcription factors involved in the regulation of the LMP 1 promoter (F~hraeus et al., 1994). On the other hand, there are several possible candidates for the targeting of EBNA2 to the promoter: RBP-Jtc at the Jtc site (Laux et al.,


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50


1994a), the PU box-binding proteins including PU.1 (Johannsen et al., 1995), the POU domain protein at the Dc~ site (this investigation) and unidentified factor(s) binding to promoter-proximal ( - 1 0 6 / + 4 0 ) sites (A. Sj6blom, A. Jansson, W. Yang & L. Rymo, unpublished results). In the present investigation, the PU box in Dfl and PU.1 could not convey EBNA2 responsiveness alone since a construct containing only Dfl [Table 1; pgLRS(-181/ -145)TKCAT] was not induced by EBNA2. This lack of response was also seen in the series of plasmids derived from p L R S ( - 1 0 6 ) ( - 1 8 1 / - 145)CAT, which contain Dfl but not D~ and showed the same level of responsiveness as the -106 LRS construct (Fig. 2), irrespective of whether Dfl was mutated. On the other hand, the POU domain protein and D~ seem to be instrumental in mediating the EBNA2 effect since a ( - 1 6 0 / - 136) LRS DNA fragment conveyed responsiveness to a basal TK promoter [Table 1; pgLRS(-160/ -136)TKCAT]. However, we have not so far been able to demonstrate a direct physical association between EBNA2 and the POU domain protein with EMSA supershift experiments (data not shown). Similar results have been reported by Sauder et al. (1994). They identified a protein-binding region in a D N A fragment of the LMP1 promoter that corresponds to the Dc~/Dfl region studied in the present investigation. No interaction between EBNA2 and the LMP 1 promoter fragment in EMSA supershift experiments could be found under conditions that clearly showed binding of EBNA2 to the LMP2A promoter. However, analysis by sucrose gradient centrifugation provided evidence that the LMP1 promoter-binding proteins form a complex that sediments with a higher velocity in EBNA2-positive cell extracts. They suggest that EBNA2-positive cells might contain specific complexes bound to the LMP1 promoter that are too labile to be detected by EMSA (Sauder et al., 1994).

The regulation of the immediate early (IE) genes by VP16 in the herpes simplex virus (HSV) system (Thomp- son & McKnight, 1992) has been used as a model to explain EBNA2 regulation of EBV promoters via RBP- JK and its binding site (Laux et al., 1994a). A similar model could be made for the interaction of EBNA2 with the POU domain protein and the PU. 1 factor and the D~ and Dfl elements. The enhancers associated with the HSV IE genes contain one or more copies of two different cis-regulatory sequences: 5' TAATGARAT 3' (' tatgarat') and the direct repeat 5' CGGAAR 3' (' cigar') motifs. The tatgarat motif forms a complex with Oct-l, VP16 and a host cell factor and the repeated cigar motif binds a tetrameric complex consisting of the subunits of an ets gene family member, the GA-binding protein. These neighbouring protein complexes then engage in protein-protein interactions with each other, creating a

functional multicomponent complex. The PU.1 transcription factor (and/or Spi-B) bound to Dfl and the POU domain protein at the octamer site in D~ and EBNA2 might interact in an analogous way.

In conclusion, our findings are consistent with the idea that efficient EBNA2 transactivation of the LMP1 promoter requires the cooperation of multiple factors that mediate their effect through different DNA sequences. We have identified a POU domain protein with the ability to target EBNA2 to a promoter. Functional cooperation between the POU domain protein and PU. 1 (and/or Spi-B) may contribute to the B cell-specific activation of the LMP1 promoter. It is possible that EBNA2 might stabilize the association between the POU domain protein, PU. 1 and EB2RE by forming a complex that cannot easily be detected by EMSA. It is also possible that post-transcriptional modifications of either PU. 1 or the POU domain protein, or both, may occur.

We thank Carina Str6m and Jane L6fvenmark for skilful technical assistance. We are very grateful to Dr Peter O'Hare for the POU domain-specific antibodies. This study was supported by grants from the Swedish MRC (project 5667), the Swedish Cancer Society, the Inga-Britt och Arne Lundberg Foundation and by National Cancer Institute Grant 28380-08. S. L. was a recipient of an EMBO Long Term Fellowship.

R e f e r e n c e s

ABBOT, S. D., ROWE, M., CADWALLADER, K., RICKSTEN, A., GORDON, J., WANG, F., RVMO, L. & RZCKINSON, A. B. (1990). Epstein-Barr virus nuclear antigen 2 expression of the virus-encoded latent membrane protein. Journal of Virology 64, 2126-2134.

BAER, R., BANKIER, m. T., BIGGIN, M. D., DEININGER, P. L., FARRELL, P. J., GIBSON, T. J., HATFULL, G., HUDSON, G. S., SATCHWELL, S. C., S~GUIN, C., TUFFNELL, P. S. & BARRELL, B. G. (1984). DNA sequence and expression of the B95-8 Epstei~Barr virus genome. Nature 310, 207-211.

BEN-BASSAT, H., GOLDBLUM, N., MITRANI, S., GOLDBLUM, T., YOFFEY, J.M., COHEN, M.M., BENTWlTCH, Z., RAMOT, B., KLEIN, E. & KLEIN, G. (1977). Establishment in continuous culture of a new type of lymphocyte from a 'Burkitt-like' malignant lymphoma (line D. G.-75). International Journal of Cancer 19, 2%33.

COHEN, J. I., WANG, F., MANNICK, J. & KIEFF, E. (1989). Epstein-Barr virus nuclear protein 2 is a key determinant of lymphocyte transformation. Proceedings' of the National Academy of Sciences', USA 86, 9558-9562.

CORDIER, M., CALENDER, A., BILLAUD, M., ZIMBER, U., ROUSSELET, G., PAVLISH, O., BANCHEREAU, J., TURSZ, T., BORNKAMM, G. & LENOIR, G .M. (1990). Stable transfection of Epstei~Barr virus (EBV) nuclear antigen 2 in lymphoma cells containing the EBV P3HR1 genome induces expression of B-cell activation molecules CD21 and CD23. Journal of Virology 64, 1002 1013.

DIGNAM, J. D., LEBOVITZ, R. M. & ROEDER, R. G. (1983). Accurate transcription initiation by RNA polymerase lI in a soluble extract from isolated mammalian nuclei. Nucleic Acids Research 11, 1475 1489.

EOLUND, T., WALKER, M. D., BARR , P. J. & RUTTER, W. J. (1985). Cell-specific expression of the rat insulin gene: evidence for role of two distinct 5' flanking elements. Science 230, 912 916.

EPSTEIN, M. A., ACHONG, B. G., BARR, Y. M., ZAJAC, B., HENLE, G. & HENLE, W. (1966). Morphological and virological investigation on cultured Burkitt tumor lymphoblasts (strain Raji). Journal of the National Cancer Institute 37, 547-559.

ERNBERG, I., FALK, K., MINAROVITS, J., BUSSON, P., TURSZ, T.,


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50


MASUCCI, M. G. & KLEIN, G. (1989). The role of methylation in the phenotype-dependent modulation of Epstein-Barr nuclear antigen 2 and latent membrane protein genes in cells latently infected with Epstein-Barr virus. Journal of General Virology 70, 2989-3002.

FAHRAEUS, R., JANSSON, A., RICKSTEN, A., SJOBLOM, A. & RYMO, L. (1990). Epstein-Barr virus-encoded nuclear antigen 2 activates the viral latent membrane protein promoter by modulating the activity of a negative regulatory element. Proceedings of the National Academy of Sciences, USA 87, 7390-7394.

F.MHRAEUS, R., JANSSON, A., SJOBLOM, A., NILSSON, T., KLEIN, G. & RYMO, L. (1993). Cell phenotype dependent control of Epstein-Barr virus latent membrane protein 1 (LMP1) gene regulatory sequences. Virology 195, 71 80.

F~IRAEUS, R., PALMQVIST, L., NERSTEDT, A., FARZAD, S., RYMO, L. & LAIN, S. (1994). Response to cAMP levels of the Epstein-Barr virus EBNA2-inducible LMP1 oncogene and EBNA2 inhibition of a PP1- like activity. EMBO Journal 13, 6041 6051.

FENNEWALD, S., VAN SANTEN, V. 8¢ KIEFF, E. (1984). Nucleotide sequence of an mRNA transcribed in latent growth-transforming virus infection indicates that it may encode a membrane protein. Journal of Virology 51, 411419.

FINKE, J., FRITZEN, R., TERNES, P., TRIVEDI, P., BROSS, K. J., LANGE, R., MERTELSMANN, R. & D6LKEN, G. (1992). Expression of bcl-2 in Burkitt's lymphoma cell lines: induction by latent Epstein Barr virus genes. Blood 80, 45%469.

GHOSH, D. & KIEFF, E. (1990). Cis-acting regulatory elements near the Epstein-Barr virus latent-infection membrane protein transcriptional start site. Journal of Virology 64, 1855-1858.

GOEBL, M. G. (1990). The PU.1 transcription factor is the product of the putative oncogene Spi-1. Cell 61, 1165-1166.

GROSSMAN, S. R., JOHANNSEN, E., TONG, X., YALAMANCHILI, R. & KIEEF, E (1994). The Epstein-Barr virus nuclear antigen 2 transactivator is directed to response elements by the JK recombination signal binding protein. Proceedings of the National Academy of Sciences, USA 91, 7568-7572.

HAMMERSCHMIDT, W. & SUGDEN, B. (1989). Genetic analysis of immortalizing functions of Epstei~Barr virus in human B lymphocytes. Nature 340, 393-397.

HEELER, M., DAMBAUGH, T. & KIEFF, E. (1981). Epstei~Barr virus DNA. IX. Variation among viral DNAs from producer and nonproducer infected cells. Journal of Virology 38, 632-648.

HENKEE, T., LIt~G, P. D., HAYWARD, S. D. & PETERSON, M. G. (1994). Mediation of Epstein-Barr virus EBNA2 transactivation by recombination signal-binding protein JK. Science 265, 92-95.

HENLE, G. & HENLE, W. (1966). Immunoftuorescence in cells derived from Burkitt's lymphoma. Journal of Bacteriology 91, 1248-1256.

HERBST, H. D. F., HUMMEL, M., NIEDOBITEK, S., PILERI, S., MULLER- LANTZSCH, N. & STEIN, H. (1991). Epstein-Barr virus latent membrane protein expression in Hodgkin and Reed-Sternberg cells. Proceedings of the National Academy of Sciences, USA 88, 4766- 4770.

JOHANNSEN, E., KOH, E., MOSIALOS, G., TONG, X., KIEFF, E. & GROSSMAN, S. (1995). Epstei~Barr virus nuclear protein 2 transactivation of the latent membrane protein 1 promoter is mediated by JK and PU. 1. Journal of Virology 69, 253-262.

KAYE, K. M., IZUMI, K.M. & KIEFF, E. (1993). EpsteimBarr virus latent protein 1 is essential for B-lymphocyte growth transformation. Proceedings of the National Academy of Sciences, USA 90, 9150- 9154.

KING, W., TOMAS-POWELL, A., RAAB-TRAUB, N., HAWKE, M. C. & KIEFF, E. (1980). Viral RNA in a restringently infected, growth- transformed cell line. Journal of Virology 36, 506-518.

KLEIN, G , DOMBOS, L. & GOTHOSKAR, B. (1972). Sensitivity of Epstein-Barr virus (EBV) producer and non-producer human lymphoblastoid cell lines to superinfection with EB-virus. Inter- national Journal of Cancer 10, 44-57.

KLEMSZ, M. J., MCKERCHER, S. R., CELADA, A., VAN BEVEREN, C. & MAKI, R. A. (1990). The macrophage and B cell-specific transcription factor PU.1 is related to the ets oncogene. Cell 61, 113 124.

KNUTSON, J. (1990). The level of c-fgr RNA is increased by EBNA-2, an Epstein-Barr virus gene required for B-cell immortalization. Journal of Virology 64, 2530-2536.

LAUX, G., DUGRILLON, F., ECKERT, C., ADAM, B., ZIMBER-STROBL, W. & BORNKAMM, G. W. (1994 a). Identification and characterization of an Epstein-Barr virus nuclear antigen 2-responsive cis element in the bidirectional promoter region of latent membrane protein and terminal protein 2 genes. Journal of Virology 68, 6947 6958.

LAUX, G., ADAM, B., STROBL, L. J. & MOKEAU-GACHEUN, F. (1994b). The Spi-1/PU.1 and Spi-B ets family transcription factors and the recombination signal binding protein RBP-Jx interacts with an Epstein-Barr virus nuclear antigen 2 responsive cis-element. EMBO Journal 13, 5624-5632.

LIEBOWlTZ, D. & Kt~FE, E. (1993). Epstein-Barr virus. In The Human Herpesviruses, pp. 10%172. Edited by B. Roizman, R. J. Whitley & C. Lopez. New York: Raven Press.

LING, P.D., RAWL1NS, D.R. & HAYWARD, S.D. (1993). The Epstein-Barr virus immortalizing protein EBNA2 is targeted to DNA by a cellular enhancer-binding protein. Proceedings of the National Academy of Sciences, USA 90, 923~9241.

LING, P. D., HSIEH, J. J.-D., RUF, I. K., RAWLINS, D. R. & HAYWARD, S. D. (1994). EBNA-2 upregulation of Epstei~Barr virus latency promoters and the cellular CD23 promoter utilizes a common targeting intermediate, CBF1. Journal of Virology 68, 5375 5383.

LONGNECKER, R., MILLER, C. L., MIAO, X.-Q, MARCHINI, m. & KIEFF, E. (1992). The only domain which distinguishes Epstein-Barr virus latent membrane protein 2A (LMP2A) from LMP2B is dispensable for lymphocyte infection and growth transformation in vitro; LMP2A is therefore nonessential. Journal of Virology 66, 646• 6469.

LONGNECKER, R., MILLER, C. L., MIAO, X.-Q., TOMKINSON, B. & KIEFF, E. (1993). The last seven transmembrane and carboxy-terminal cytoplasmic domains of Epstei~Barr virus latent membrane protein 2 (LMP2) are dispensible for lymphocyte infection and growth transformation in vitro. Journal of Virology 67, 2006-2013.

MANNICK, J. B., COHEN, J. I., BIRKENBACH, M., MARCHINI, A. & KIEFF, E. (1991). The Epstein-Barr virus nuclear protein encoded by the leader of the EBNA RNAs is important in B-lymphocyte transformation. Journal of Virology 65, 6826-6837.

MEITINGER, C., STROBE, t . J., MARSCHALL, G., BORNKAMM, G. W. ZIMBER-STROBL, U. (1994). Crucial sequences within the Epstei~ Barr virus TP1 promoter for EBNA2-mediated transactivation and interaction of EBNA2 with its responsive element. Journal of Virology 68, 7497 7506.

MILLER, G. (1990). Epstein-Barr virus: biology, pathogenesis and medical aspects. In Virology, 2nd edn, pp. 1921-1958. Edited by B. N. Fields and D. M. Knipe. New York: Raven Press.

OFVERSTEDT, t . G., HAMMARSTROM, K., BALGOBIN, N., HJERTI~N, S., PETTERSSON, U. O. CHATTOPADHYAYA, J. (1984). Rapid and quan- titative recovery of DNA fragments from gels by displacement electrophoresis. Biochimica et Biophysica Acta 782, 120-126.

RAY, D., BOSSELUT, J., GHYSDAEL, J., MATTEI, M.-G., TAVITIAN, A. ~¢ MOREAU-GACHELIN, F. (1992). Characterization of Spi-B, a transcription factor related to the putative oncoprotein Spi-1/PU.1. Molecular and Cellular Biology 12, 4297-4304.

RICKSTEN, A., SVENSSON, C., WELINDER, C. 8¢ RYMO, L. (1987). Identification of sequences in Epstein-Barr virus DNA required for the expression of the second Epstein-Barr virus-determined nuclear antigen in COS-1 cells. Journal of General Virology 68, 2407 2418.

RICKSTEN, A., OLSSON, A., ANDERSSON, T. &. RYMO, L. (1988). The 5' flanking region of the gene for the Epstei~Barr virus-encoded nuclear antigen 2 contains a cell type specific cis-acting regulatory element that activates transcription in transfected B-cells. Nucleic Acids Research 16, 8391-8410.

SADOWSKI, H. B., SHUAI, K., DARNELL, J.E., JR & GILMAN, M.Z. (1993). A common nuclear signal transduction pathway activated by growth factor and cytokine receptors. Science 261, 1739 1744.

SAUDER, C., HAISS, P., GR.~SSER, F. A., ZIMBER-STROBL, U. ~,~ MUELLER- LANTZSCH, N. (1994). DNA-binding studies of the Epstei~Barr virus nuclear antigen 2 (EBNA-2): evidence for complex formation by latent membrane protein gene promoter-binding proteins in EBNA-2-positive cell lines. Journal of General Virology 75, 306~ 3079.

SJOBLOM, A., JANSSON, A., NILSSON, T., LAIN, S. & RYMO, L. (1993). On the mechanism of transcriptional activation of the LMP1 promoter by EBNA2. In The Epstein-Barr Virus and Associated Diseases, pp.


IP: 54.145.26.59

On: Fri, 04 Mar 2016 19:34:50


153 158. Edited by T. Tursz, J. S. Pagano, D. V. Ablashi, G. de Th6, G. Lenoir & G. R. Pearson. Paris & London: John Libbey Eurotext.

SJ6BLOM, A., NERSTEDT, A., JANSSON, A. & RYMO, L. (1995). Domains of the Epstein-Barr virus nuclear antigen 2 (EBNA2) involved in the transactivation of the latent membrane protein l and the EBNA Cp promoters. Journal of General Virology 76, 2669-2678.

SKARE, J., FARLEY, J., STROMINGER, J. L., FRESEN, K. O., CHO, M. S. & ZUR HAUSEN, H. (1985). Transformation by Epstein-Barr virus requires DNA sequences in the region of BamHI fragments Y and H. Journal of Virology 55, 28~%297.

SUNG, M. S., KENNEY, S., GUTSCH, B. & PAGANO, J. S. (1991). EBNA- 2 transactivates a lymphoid-specific enhancer in the BamHI C promoter of Epstei~Barr virus. Journal of Virology 65, 2164-2169.

SWAMINATHAN, S., TOMKINSON, B. & KIEEF, E. (1991). Recombinant Epstei~Barr virus with small RNA (EBER) genes deleted trans- forms lymphocytes and replicates in vitro. Proceedings of the National Academy of Sciences, USA 88, 1546-1550.

THOMPSON, C. C. & McKNIGHT, S. L. (1992). Anatomy of an enhancer. Trends in Genetics 8, 232-236.

TOMKINSON, B. & KIEFE, E. (1992). Use of second-site homologous recombination to demonstrate that Epstein-BarE virus nuclear protein 3B is not important for lymphocyte infection or growth transformation in vitro. Journal of Virology 66, 2893-2903.

TOMKINSON, B., ROBERTSON, E. & KIEFF, E. (1993). Epstein-BarE virus nuclear proteins EBNA-3A and EBNA-3C are essential for B- lymphocyte growth transformation. Journal of Virology 67, 2014- 2025.

TUN, T., HAMAGUCHI, Y., MATSUNAMI, N., FURUKAWA, T., HONJO, T. & KAWAICHI, M. (1994). Recognition sequence of a highly conserved DNA binding protein RBP-JK. Nucleic Acids Research 22, 965-971.

WALTZER, L., LOGEAT, F., BROU, C., ISRAEL, A., SERGEANT, A. & MANET, E. (1994). The human J• recombination signal sequence binding protein (RBP-Jx) targets the Epstein-Barr virus EBNA2 protein to its DNA responsive elements. EMBO Journal 13, 5633-5638.

WANG, F., GREGORY, C. D., ROWE, M., RICK1NSON, A. B., WANG, D., BIRKENBACH, M., KIKUTANI, H., KISHIMOTO, T. & K1EEF, E. (1987). Epstein-Barr virus nuclear antigen 2 specifically induces expression of the B-cell activation antigen CD23. Proceedings of the National Academy of Sciences, USA 84, 3452-3456.

WANG, F., TSANG, F.-S., KURILLA, M. G., COHEN, J. I. & KIEFF, E. (1990a). Epstein-Barr virus nuclear antigen 2 transactivates latent membrane protein LMP1. Journal of Virology 64, 3407-3416.

WANG, F., GREGORY, C., SAMPLE, C., Rowe, M., LIEBOWITZ, D., MURRAY, R., RICKINSON, A. & KIEFF, E. (1990b). Epstein-Barr virus latent membrane protein (LMP1) and nuclear proteins 2 and 3C are effectors of phenotypic changes in B lymphocytes: EBNA-2 and LMP1 cooperatively induce CD23. Journal of Virology 64, 2309- 2318.

WANG, F., KIKUTANI, H., TSANG, S.-F., KISHIMOTO, T. & KIEFF, E. (1991). Epstein Barr virus nuclear protein 2 transactivates a cis- acting CD23 DNA element. Journal of Virology 65, 4101~,106.

WEGNER, M., DROLET, D.W. & ROSENFELD, M.G. (1993). POU domain proteins: structure and function of developmental regu- lators. Current Opinion in Cell Biology 5, 488~98.

WOISETSCHLAEGER, M., YANDAVA, C.N., FURMANSKI, L.A., STRO- MINGER, J.L. & SPECK, S.H. (1990). Promoter switching in Epstei~Barr virus during the initial stages of infection of B lymphocytes. Proceedings of the National Academy of Sciences, USA 87, 1725-1729.

WOISETSCHLAEGER, M., JIN, X.W., YANDAVA, C.N., FURMANSKI, L.A., STROMINGER, J.L. & SPECK, S.H. (1991). Role for the Epstei~Barr virus nuclear antigen 2 in viral promoter switching during initial stages of infection. Proceedings of the National Academy of Sciences, USA 88, 3942-3946.

YALAMANCHILI, R., TONG, X., GROSSMAN, S. E. J., MOSIALOS, G. & KIEFF, E. (1994). Genetic and biochemical evidence that EBNA2 interaction with a 63-kDa cellular GTG-binding protein is essential for B lymphocyte growth transformation by EBV. Virology 204, 634-641.

YATES, J. L., WARREN, N. & SUGDEN, B. (1985). Stable replication of plasmids derived from Epstein-Barr virus in a variety of mammalian cells. Nature 313, 81~815.

YOUNG, L. S., ALFIERI, C., HENNESSY, K., EVANS, H., O'HARA, C., ANDERSSON, H. G., RITZ, J., SHAPIRO, S., RICKINSON, A., KIEFF, E. & COHEN, J.I. (1989). Expression of Epstein Barr virus transformation-associated genes in tissues of patients with EBV lymphoproliferative disease. New England Journal of Medicine 321, 108~ 1085.

ZIMBER-STROBL, U., KREMMER, E., GR.g, SSER, E., MARSCHALL, G., LAUX, G. & BORNKAMM, G.W. (1993). The Epstei~Barr virus nuclear antigen 2 interacts with an EBNA2 responsive cis-element of the terminal protein 1 gene promoter. EMBO Journal 12, 167 175.

(Received 29 March 1995; Accepted 9 June 1995)

PU box-binding transcription factors and a POU domain protein cooperate in the Epstein--Barr virus...

Documents

Transcript of PU box-binding transcription factors and a POU domain protein cooperate in the Epstein--Barr virus...